import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import plotly.express as px
from matplotlib.colors import LinearSegmentedColormap
import matplotlib.patches as patches
import seaborn as sns
import warnings
warnings.filterwarnings("ignore")

# To show all columns when printing
pd.set_option("display.max_columns", None)

df = pd.read_csv(r"/Users/maniisshhhh/Downloads/Healthcare.csv")
df

	Age	Gender	BMI	Glucose	BloodPressure	Insulin	SkinThickness	Pregnancies	DiabetesPedigreeFunction	PhysicalActivityLevel	Outcome	Country	Cholesterol	Smoking_Status	Family_History	Diet_Type	Sleep_Hours	Daily_Steps	Water_Intake_Liters	Mental_Health_Status	Sleep_Quality	Chronic_Disease	Work_Stress_Level	Alcohol_Consumption	Diabetes_Status	Diabetes_Type	Fast_Food_Intake	Prediabetes	Fasting_Blood_Sugar	Screen_Time
0	56	Female	27.0	76.5	68.90	16.40	8.9	2	1.181	Moderate	0	Australia	189.8	Yes	No	High-Fat	8.5	4878.0	3.7	Good	Fair	Thyroid	High	Regular	No Diabetes	Type 2	Rarely	No	185.4	8.9
1	69	Male	28.9	43.5	66.20	31.90	20.5	2	0.339	Moderate	0	UK	235.4	Yes	Yes	High-Carb	4.9	5604.0	3.1	Good	Fair	Heart Disease	Moderate	Occasional	No Diabetes	Type 2	Rarely	Yes	231.7	1.9
2	46	Female	27.4	82.1	75.25	79.10	16.1	2	1.659	High	1	South Africa	143.7	No	No	Balanced	7.0	7533.0	2.6	Good	Good	Thyroid	Moderate	Regular	Diabetes	Type 2	Sometimes	No	64.8	10.7
3	32	Female	28.9	105.5	75.10	189.80	26.2	7	2.497	Moderate	0	Canada	219.5	Yes	No	Balanced	6.2	5963.0	2.6	Poor	Fair	Hypertension	Low	Occasional	No Diabetes	Type 2	Frequently	No	89.1	6.3
4	60	Male	24.0	84.2	79.00	100.40	24.2	8	1.342	Low	1	UK	219.0	No	Yes	High-Fat	8.0	3201.0	2.6	Average	Good	Thyroid	Moderate	Regular	Diabetes	Type 2	Frequently	No	137.7	4.7
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
17995	46	Female	32.3	84.3	86.10	59.50	15.3	6	1.019	High	0	Germany	179.8	No	Yes	Balanced	6.0	5513.0	1.2	Good	Poor	Hypertension	Low	Occasional	No Diabetes	Type 2	Sometimes	No	111.2	6.6
17996	31	Male	28.9	137.2	79.80	87.30	20.0	5	0.329	Low	1	South Africa	101.5	No	No	High-Carb	6.3	5859.0	3.4	Poor	Good	Thyroid	Low	Regular	Diabetes	Type 2	Rarely	Yes	77.4	6.1
17997	21	Male	26.9	108.4	69.90	79.85	32.8	9	0.989	High	1	Germany	229.8	No	No	Balanced	5.0	5386.0	1.7	Average	Fair	Thyroid	High	Occasional	Diabetes	Type 2	Sometimes	Yes	52.0	10.9
17998	61	Male	28.6	139.4	66.30	92.00	20.0	8	1.456	Moderate	0	Canada	214.1	Yes	Yes	High-Carb	8.2	12200.0	4.1	Poor	Fair	Hypertension	Moderate	Occasional	No Diabetes	Type 2	Rarely	No	171.3	9.5
17999	73	Male	28.0	135.4	77.80	115.60	21.3	6	1.398	Moderate	0	Brazil	221.1	No	Yes	High-Fat	7.3	1208.0	3.3	Average	Poor	Heart Disease	Low	Occasional	No Diabetes	Type 2	Rarely	No	141.8	1.9

18000 rows × 30 columns

Basic Dataset Overview

print("Shape of dataset:", df.shape)

Shape of dataset: (18000, 30)

print("\nFirst 5 rows:")
print(df.head())

First 5 rows:
   Age  Gender   BMI  Glucose  BloodPressure  Insulin  SkinThickness  \
0   56  Female  27.0     76.5          68.90     16.4            8.9   
1   69    Male  28.9     43.5          66.20     31.9           20.5   
2   46  Female  27.4     82.1          75.25     79.1           16.1   
3   32  Female  28.9    105.5          75.10    189.8           26.2   
4   60    Male  24.0     84.2          79.00    100.4           24.2   

   Pregnancies  DiabetesPedigreeFunction PhysicalActivityLevel  Outcome  \
0            2                     1.181              Moderate        0   
1            2                     0.339              Moderate        0   
2            2                     1.659                  High        1   
3            7                     2.497              Moderate        0   
4            8                     1.342                   Low        1   

        Country  Cholesterol Smoking_Status Family_History  Diet_Type  \
0     Australia        189.8            Yes             No   High-Fat   
1            UK        235.4            Yes            Yes  High-Carb   
2  South Africa        143.7             No             No   Balanced   
3        Canada        219.5            Yes             No   Balanced   
4            UK        219.0             No            Yes   High-Fat   

   Sleep_Hours  Daily_Steps  Water_Intake_Liters Mental_Health_Status  \
0          8.5       4878.0                  3.7                 Good   
1          4.9       5604.0                  3.1                 Good   
2          7.0       7533.0                  2.6                 Good   
3          6.2       5963.0                  2.6                 Poor   
4          8.0       3201.0                  2.6              Average   

  Sleep_Quality Chronic_Disease Work_Stress_Level Alcohol_Consumption  \
0          Fair         Thyroid              High             Regular   
1          Fair   Heart Disease          Moderate          Occasional   
2          Good         Thyroid          Moderate             Regular   
3          Fair    Hypertension               Low          Occasional   
4          Good         Thyroid          Moderate             Regular   

  Diabetes_Status Diabetes_Type Fast_Food_Intake Prediabetes  \
0     No Diabetes        Type 2           Rarely          No   
1     No Diabetes        Type 2           Rarely         Yes   
2        Diabetes        Type 2        Sometimes          No   
3     No Diabetes        Type 2       Frequently          No   
4        Diabetes        Type 2       Frequently          No   

   Fasting_Blood_Sugar  Screen_Time  
0                185.4          8.9  
1                231.7          1.9  
2                 64.8         10.7  
3                 89.1          6.3  
4                137.7          4.7  

df.columns

Index(['Age', 'Gender', 'BMI', 'Glucose', 'BloodPressure', 'Insulin',
       'SkinThickness', 'Pregnancies', 'DiabetesPedigreeFunction',
       'PhysicalActivityLevel', 'Outcome', 'Country', 'Cholesterol',
       'Smoking_Status', 'Family_History', 'Diet_Type', 'Sleep_Hours',
       'Daily_Steps', 'Water_Intake_Liters', 'Mental_Health_Status',
       'Sleep_Quality', 'Chronic_Disease', 'Work_Stress_Level',
       'Alcohol_Consumption', 'Diabetes_Status', 'Diabetes_Type',
       'Fast_Food_Intake', 'Prediabetes', 'Fasting_Blood_Sugar',
       'Screen_Time'],
      dtype='object')

df.dtypes

Age                           int64
Gender                       object
BMI                         float64
Glucose                     float64
BloodPressure               float64
Insulin                     float64
SkinThickness               float64
Pregnancies                   int64
DiabetesPedigreeFunction    float64
PhysicalActivityLevel        object
Outcome                       int64
Country                      object
Cholesterol                 float64
Smoking_Status               object
Family_History               object
Diet_Type                    object
Sleep_Hours                 float64
Daily_Steps                 float64
Water_Intake_Liters         float64
Mental_Health_Status         object
Sleep_Quality                object
Chronic_Disease              object
Work_Stress_Level            object
Alcohol_Consumption          object
Diabetes_Status              object
Diabetes_Type                object
Fast_Food_Intake             object
Prediabetes                  object
Fasting_Blood_Sugar         float64
Screen_Time                 float64
dtype: object

Check Missing Values

missing = df.isnull().sum()
missing_percent = (missing / len(df)) * 100

missing_summary = pd.DataFrame({
    "Missing Values": missing,
    "Missing %": missing_percent.round(2)
})

print("\nMissing Values Summary:")
print(
    missing_summary[missing_summary["Missing Values"] > 0]
    .sort_values("Missing %", ascending=False)
)

Missing Values Summary:
                     Missing Values  Missing %
BMI                               5       0.03
Glucose                           5       0.03
Sleep_Hours                       5       0.03
Daily_Steps                       5       0.03
Water_Intake_Liters               5       0.03

plt.figure(figsize=(12,5))
sns.heatmap(df.isnull(), cbar=False)
plt.title("Missing Data Heatmap")
plt.show()

print("\nSummary Statistics (Numeric Columns):")
df.describe()

Summary Statistics (Numeric Columns):

	Age	BMI	Glucose	BloodPressure	Insulin	SkinThickness	Pregnancies	DiabetesPedigreeFunction	Outcome	Cholesterol	Sleep_Hours	Daily_Steps	Water_Intake_Liters	Fasting_Blood_Sugar	Screen_Time
count	18000.000000	17995.000000	17995.000000	18000.000000	18000.000000	18000.000000	18000.000000	18000.000000	18000.000000	18000.000000	17995.000000	17995.000000	17995.000000	18000.000000	18000.000000
mean	48.591056	28.015793	109.903351	75.112711	80.071211	19.982994	4.448167	1.301701	0.286389	199.731939	7.016605	6020.197888	2.500072	110.207017	6.524706
std	17.629407	5.899675	34.562562	11.836675	44.434725	7.871159	2.828718	0.681769	0.452086	39.580036	1.493874	2475.720091	0.797083	35.141720	3.177019
min	18.000000	4.900000	-46.300000	25.100000	-120.800000	-9.600000	0.000000	0.100000	0.000000	33.100000	1.300000	-3980.000000	-0.600000	-18.500000	1.000000
25%	34.000000	24.100000	87.100000	67.200000	51.175000	14.800000	2.000000	0.721000	0.000000	173.400000	6.000000	4361.500000	2.000000	86.200000	3.800000
50%	48.000000	28.200000	108.700000	75.250000	79.850000	20.000000	4.000000	1.294000	0.000000	199.800000	7.000000	6060.000000	2.500000	110.200000	6.500000
75%	64.000000	32.000000	132.900000	83.000000	108.800000	25.100000	7.000000	1.884000	1.000000	226.400000	8.000000	7652.500000	3.000000	134.300000	9.300000
max	79.000000	54.900000	240.500000	119.900000	249.200000	50.600000	9.000000	2.500000	1.000000	358.300000	12.200000	15152.000000	6.100000	242.800000	12.000000

Data Type Cleaning Summary

num_cols = df.select_dtypes(include=["int64","float64"]).columns
cat_cols = df.select_dtypes(include="object").columns

# Fill numeric with median
for col in num_cols:
    df[col].fillna(df[col].median(), inplace=True)

# Fill categorical with mode
for col in cat_cols:
    df[col].fillna(df[col].mode()[0], inplace=True)

print("Missing values after filling:\n", df.isnull().sum().sum())

Missing values after filling:
 0

categorical_cols = df.select_dtypes(include="object").columns
print("\nCategorical Columns:", list(categorical_cols))

# Explore categorical columns
for col in categorical_cols:
    print(f"\nColumn: {col}")
    print("Unique values:", df[col].nunique())
    print(df[col].value_counts().head())

Categorical Columns: ['Gender', 'PhysicalActivityLevel', 'Country', 'Smoking_Status', 'Family_History', 'Diet_Type', 'Mental_Health_Status', 'Sleep_Quality', 'Chronic_Disease', 'Work_Stress_Level', 'Alcohol_Consumption', 'Diabetes_Status', 'Diabetes_Type', 'Fast_Food_Intake', 'Prediabetes']

Column: Gender
Unique values: 3
Gender
Female    9374
Male      8621
Other        5
Name: count, dtype: int64

Column: PhysicalActivityLevel
Unique values: 3
PhysicalActivityLevel
High        6383
Low         5811
Moderate    5806
Name: count, dtype: int64

Column: Country
Unique values: 8
Country
South Africa    2630
Brazil          2259
UK              2253
India           2205
Australia       2182
Name: count, dtype: int64

Column: Smoking_Status
Unique values: 3
Smoking_Status
No         13546
Yes         4449
unknown        5
Name: count, dtype: int64

Column: Family_History
Unique values: 2
Family_History
No     10891
Yes     7109
Name: count, dtype: int64

Column: Diet_Type
Unique values: 3
Diet_Type
High-Carb    6296
Balanced     5865
High-Fat     5839
Name: count, dtype: int64

Column: Mental_Health_Status
Unique values: 3
Mental_Health_Status
Good       9190
Average    5268
Poor       3542
Name: count, dtype: int64

Column: Sleep_Quality
Unique values: 4
Sleep_Quality
Good         4883
Excellent    4426
Poor         4368
Fair         4323
Name: count, dtype: int64

Column: Chronic_Disease
Unique values: 3
Chronic_Disease
Thyroid          9256
Heart Disease    4477
Hypertension     4267
Name: count, dtype: int64

Column: Work_Stress_Level
Unique values: 3
Work_Stress_Level
Low         6319
High        5879
Moderate    5802
Name: count, dtype: int64

Column: Alcohol_Consumption
Unique values: 2
Alcohol_Consumption
Regular       12129
Occasional     5871
Name: count, dtype: int64

Column: Diabetes_Status
Unique values: 2
Diabetes_Status
No Diabetes    12845
Diabetes        5155
Name: count, dtype: int64

Column: Diabetes_Type
Unique values: 2
Diabetes_Type
Type 2    17215
Type 1      785
Name: count, dtype: int64

Column: Fast_Food_Intake
Unique values: 4
Fast_Food_Intake
Rarely        7192
Sometimes     6341
Frequently    3570
Daily          897
Name: count, dtype: int64

Column: Prediabetes
Unique values: 2
Prediabetes
No     12236
Yes     5764
Name: count, dtype: int64

# Convert numeric-looking object columns to numeric
numeric_like_cols = ["BMI","Glucose","Sleep_Hours","Daily_Steps","Water_Intake_Liters","Fast_Food_Intake"]

for col in numeric_like_cols:
    df[col] = pd.to_numeric(df[col], errors="coerce")

Outlier Check (Summary)

plt.figure(figsize=(16, 20))

for i, col in enumerate(num_cols, 1):
    plt.subplot(len(num_cols)//3 + 1, 3, i)
    sns.boxplot(data=df, y=col, color="skyblue")
    plt.title(col)

plt.tight_layout()
plt.show()

plt.figure(figsize=(18, 6))

cols = ["BMI","Glucose","Sleep_Hours","Daily_Steps","Water_Intake_Liters","Fast_Food_Intake"]
sns.boxplot(data=df[cols])
plt.title("Boxplot Overview of Key Numeric Variables")
plt.show()

EDA

plt.figure(figsize=(7, 5))
sns.boxplot(data=df, x="Diabetes_Status", y="BloodPressure", palette="coolwarm")
plt.title("Blood Pressure by Diabetes Status", fontsize=14)
plt.xlabel("Diabetes Status")
plt.ylabel("Blood Pressure")
plt.tight_layout()
plt.show()

Average Cholesterol by Gender (Bar Plot)

chol_gender = df.groupby("Gender", as_index=False)["Cholesterol"].mean()

plt.figure(figsize=(7, 5))
sns.barplot(data=chol_gender, x="Gender", y="Cholesterol", palette="viridis")
plt.title("Average Cholesterol by Gender", fontsize=14)
plt.ylabel("Cholesterol")

for i, v in enumerate(chol_gender["Cholesterol"]):
    plt.text(i, v + 1, f"{v:.1f}", ha="center")
    
plt.tight_layout()
plt.show()

Daily Steps vs BMI (Scatter Plot)

sample_df = df.sample(500, random_state=42)

plt.figure(figsize=(7, 5))
sns.scatterplot(data=sample_df,x="Daily_Steps",y="BMI",hue="Outcome",palette="coolwarm",alpha=0.6)
plt.title("Daily Steps vs BMI (Random Sample of 500 Rows)", fontsize=14)
plt.xlabel("Daily Steps")
plt.ylabel("BMI")
plt.tight_layout()
plt.show()

Random Sample (2000 rows)

sample_df = df.sample(2000, random_state=42)

plt.figure(figsize=(7, 5))
sns.scatterplot(data=sample_df,x="Daily_Steps",y="BMI",hue="Outcome",palette="coolwarm",alpha=0.6)
plt.title("Daily Steps vs BMI (Random Sample of 2000 Rows)", fontsize=14)
plt.xlabel("Daily Steps")
plt.ylabel("BMI")
plt.tight_layout()
plt.show()

More Detailed

plt.figure(figsize=(7, 5))
sns.scatterplot(data=df, x="Daily_Steps", y="BMI", hue="Outcome", palette="coolwarm", alpha=0.6)
plt.title("Daily Steps vs BMI", fontsize=14)
plt.xlabel("Daily Steps")
plt.ylabel("BMI")
plt.tight_layout()
plt.show()

Sleep Hours Distribution (Histogram)

plt.figure(figsize=(7, 5))
sns.histplot(df["Sleep_Hours"], kde=True, bins=30, color="purple")
plt.title("Distribution of Sleep Hours", fontsize=14)
plt.xlabel("Sleep Hours")
plt.tight_layout()
plt.show()

Categorical Frequency Heatmap

cat_cols = df.select_dtypes(include="object").columns

freq_df = df[cat_cols].apply(lambda x: x.value_counts())
plt.figure(figsize=(12,7))
sns.heatmap(freq_df, cmap="viridis")
plt.title("Categorical Feature Frequency Heatmap")
plt.tight_layout()
plt.show()

BMI by Diet Type (Box Plot)

plt.figure(figsize=(8,5))
sns.boxplot(data=df, x="Diet_Type", y="BMI", palette="magma")
plt.title("BMI by Diet Type")
plt.tight_layout()
plt.show()

UNIVARIATE ANALYSIS (Single variable)

Numeric Variables (Histograms)

df[num_cols].hist(figsize=(15,10), bins=30)
plt.suptitle("Distribution of Numerical Features")
plt.show()

Categorical Variables (Countplots)

import math

n = len(cat_cols)
rows = math.ceil(n / 3)

plt.figure(figsize=(20,20))
for i, col in enumerate(cat_cols, 1):
    plt.subplot(rows, 3, i)
    sns.countplot(data=df, x=col, palette="viridis")
    plt.title(col)

plt.tight_layout()
plt.show()

Boxplots for Outlier Detection

# Select all numeric columns
num_cols = df.select_dtypes(include=["float64", "int64"]).columns

plt.figure(figsize=(18, 6))
sns.boxplot(data=df[num_cols], palette="Set3")
plt.title("Boxplot Overview of All Numeric Features", fontsize=14)
plt.tight_layout()
plt.show()

BIVARIATE ANALYSIS (Two variables)

plt.figure(figsize=(8, 5))
sns.boxplot(data=df, x="Mental_Health_Status", y="Sleep_Hours", palette="viridis")

plt.title("Sleep Hours by Mental Health Status", fontsize=14)
plt.xlabel("Mental Health Status", fontsize=12)
plt.ylabel("Sleep Hours", fontsize=12)
plt.tight_layout()
plt.show()

Average BMI by Smoking Status

bmi_smoking = (df.groupby("Smoking_Status", as_index=False)["BMI"].mean().sort_values("BMI", ascending=False))

plt.figure(figsize=(10, 5))
sns.barplot(data=bmi_smoking, x="Smoking_Status", y="BMI", palette="viridis")

for index, row in bmi_smoking.iterrows():
    plt.text(index, row["BMI"] + 0.2, round(row["BMI"], 1),ha='center', fontsize=10)

plt.title("Average BMI by Smoking Status", fontsize=14)
plt.xlabel("Smoking Status")
plt.ylabel("Average BMI")
plt.tight_layout()
plt.show()

sns.boxplot(data=df, x="Diabetes_Type", y="Age", palette="coolwarm")
plt.title("Age by Diabetes Type")
plt.show()

Random Sample of 500 Rows

sample_df = df.sample(500, random_state=42)

plt.figure(figsize=(6,4))
sns.scatterplot(data=sample_df,x="BMI",y="Glucose",hue="Gender",alpha=0.6)
plt.title("BMI vs Glucose (Random Sample of 500 Rows)")
plt.tight_layout()
plt.show()

Random Sample of 2000 Rows

sample_df = df.sample(2000, random_state=42)

plt.figure(figsize=(6,4))
sns.scatterplot(data=sample_df,x="BMI",y="Glucose",hue="Gender",alpha=0.6)
plt.title("BMI vs Glucose (Random Sample of 2000 Rows)")
plt.tight_layout()
plt.show()

More Detailed

plt.figure(figsize=(6,4))
sns.scatterplot(data=df, x="BMI", y="Glucose", hue="Gender", alpha=0.6)
plt.title("BMI vs Glucose")
plt.tight_layout()
plt.show()

Glucose Levels by Gender

# Average Glucose Level by Gender
glucose_by_gender = (df.groupby("Gender", as_index=False)["Glucose"].mean().sort_values("Glucose", ascending=False))

plt.figure(figsize=(10, 5))
sns.barplot(data=glucose_by_gender, x="Gender", y="Glucose", palette="viridis")
plt.title("Average Glucose Level by Gender", fontsize=14)
plt.xlabel("Gender", fontsize=12)
plt.ylabel("Average Glucose Level", fontsize=12)

for index, row in glucose_by_gender.iterrows():
    plt.text(index, row["Glucose"] + 1, round(row["Glucose"], 1),ha='center', fontsize=10)

plt.tight_layout()
plt.show()

plt.figure(figsize=(8,5))
sns.boxplot(data=df, x="Gender", y="Glucose", palette="coolwarm")
plt.title("Glucose by Gender")
plt.tight_layout()
plt.show()

Diabetes Outcome vs Age

plt.figure(figsize=(8,5))
sns.histplot(data=df, x="Age", hue="Diabetes_Status", kde=True, palette="coolwarm")
plt.title("Age Distribution by Diabetes Status")
plt.xlabel("Age")
plt.tight_layout()
plt.show()

BMI by Smoking Status

plt.figure(figsize=(10,5))
sns.boxplot(data=df, x="Smoking_Status", y="BMI", palette="Set2")
plt.title("BMI by Smoking Status")
plt.tight_layout()
plt.show()

Random Sample of 500 Rows

sample_df = df.sample(500, random_state=42)

plt.figure(figsize=(6, 4))
sns.scatterplot(data=sample_df,x="Glucose", y="Insulin",alpha=0.7, edgecolor=None, color="teal")
plt.title("Glucose vs Insulin (Sample of 500 Rows)", fontsize=14)
plt.xlabel("Glucose Level", fontsize=12)
plt.ylabel("Insulin Level", fontsize=12)
plt.tight_layout()
plt.show()

Random Sample of 2500 Rows

sample_df = df.sample(2500, random_state=42)

plt.figure(figsize=(6, 4))
sns.scatterplot(data=sample_df,x="Glucose", y="Insulin",alpha=0.7, edgecolor=None, color="teal")
plt.title("Glucose vs Insulin (Sample of 2500 Rows)", fontsize=14)
plt.xlabel("Glucose Level", fontsize=12)
plt.ylabel("Insulin Level", fontsize=12)
plt.tight_layout()
plt.show()

More Detailed

plt.figure(figsize=(6, 4))
sns.scatterplot(data=df,x="Glucose", y="Insulin",alpha=0.7, edgecolor=None, color="teal")

plt.title("Glucose vs Insulin", fontsize=14)
plt.xlabel("Glucose Level", fontsize=12)
plt.ylabel("Insulin Level", fontsize=12)
plt.tight_layout()
plt.show()

Compares Glucose vs Diabetes Outcome

plt.figure(figsize=(6, 4))
sns.boxplot(data=df, x="Diabetes_Status", y="Glucose", palette="coolwarm")
plt.title("Glucose Levels by Diabetes Status", fontsize=14)
plt.xlabel("Diabetes Status", fontsize=12)
plt.ylabel("Glucose Level", fontsize=12)
plt.tight_layout()
plt.show()

Relationship between Age and BMI (Regplot)

Random Sample of 500

sample_500 = df.sample(500, random_state=42)

plt.figure(figsize=(6, 4))
sns.regplot(data=sample_500, x="Age", y="BMI", scatter_kws={"alpha":0.6}, line_kws={"color":"red"})
plt.title("Age vs BMI (Sample of 500)", fontsize=14)
plt.xlabel("Age", fontsize=12)
plt.ylabel("BMI", fontsize=12)
plt.tight_layout()
plt.show()

Random Sample of 2500

sample_2500 = df.sample(2500, random_state=42)

plt.figure(figsize=(6, 4))
sns.regplot(data=sample_2500, x="Age", y="BMI",scatter_kws={"alpha":0.6}, line_kws={"color":"red"})
plt.title("Age vs BMI (Sample of 2500)", fontsize=14)
plt.xlabel("Age", fontsize=12)
plt.ylabel("BMI", fontsize=12)
plt.tight_layout()
plt.show()

MULTIVARIATE ANALYSIS (Three or more variables)

Correlation Heatmap (All numeric variables)

plt.figure(figsize=(14,10))
sns.heatmap(df.corr(numeric_only=True), annot=True, cmap="coolwarm")
plt.title("Correlation Heatmap")
plt.show()

Pairplot of Top Health Indicators

with Random Sample of 500 Rows

sample_500 = df.sample(500, random_state=42)
sns.pairplot(sample_500[["Age", "BMI", "Glucose", "Insulin", "Cholesterol", "Diabetes_Status"]],hue="Diabetes_Status",diag_kind="kde")
plt.show()

sns.pairplot(df[["Age", "BMI", "Glucose", "Insulin", "Cholesterol", "Diabetes_Status"]],
             hue="Diabetes_Status", diag_kind="kde")
plt.show()

Water Intake vs BP colored by Diabetes Outcome

Random Sample of 500 Rows

sample_df = df.sample(500, random_state=42)

plt.figure(figsize=(8,5))
sns.scatterplot(data=sample_df,x="Water_Intake_Liters",y="BloodPressure",hue="Diabetes_Status",alpha=0.7,palette="viridis")
plt.title("Water Intake vs Blood Pressure (Sample of 500 Rows)")
plt.tight_layout()
plt.show()

Random Sample of 1500 Rows

sample_df = df.sample(1500, random_state=42)

plt.figure(figsize=(8,5))
sns.scatterplot(data=sample_df,x="Water_Intake_Liters",y="BloodPressure",hue="Diabetes_Status",alpha=0.7,palette="viridis")
plt.title("Water Intake vs Blood Pressure (Sample of 1500 Rows)")
plt.tight_layout()
plt.show()

More Detailed

plt.figure(figsize=(8,5))
sns.scatterplot(data=df, x="Water_Intake_Liters", y="BloodPressure",hue="Diabetes_Status", alpha=0.6, palette="viridis")
plt.title("Water Intake vs Blood Pressure")
plt.tight_layout()
plt.show()

Diet Type + BMI + Diabetes Outcome (Boxplot)

plt.figure(figsize=(10,5))
sns.boxplot(data=df, x="Diet_Type", y="BMI", hue="Diabetes_Status", palette="viridis")
plt.title("Diet Type vs BMI vs Diabetes_Status")
plt.tight_layout()
plt.show()

Glucose vs Insulin

Random Sample of 500 Rows

sample_500 = df.sample(500, random_state=42)

plt.figure(figsize=(6,4))
sns.scatterplot(data=sample_500,x="Glucose", y="Insulin", hue="Diabetes_Status", size="Age", alpha=0.6, palette="viridis")
plt.title("Glucose vs Insulin (Sample 500 | Colored by Diabetes_Status, Sized by Age)")
plt.tight_layout()
plt.show()

Random Sample of 2500 Rows

sample_500 = df.sample(2500, random_state=42)

plt.figure(figsize=(6,4))
sns.scatterplot(data=sample_500,x="Glucose", y="Insulin", hue="Diabetes_Status", size="Age", alpha=0.6, palette="viridis")
plt.title("Glucose vs Insulin (Sample 2500 | Colored by Diabetes_Status, Sized by Age)")
plt.tight_layout()
plt.show()

More Detailed

plt.figure(figsize=(6,4)) 
sns.scatterplot(data=df, x="Glucose", y="Insulin", hue="Diabetes_Status", size="Age", alpha=0.6, palette="viridis") 
plt.title("Glucose vs Insulin (Colored by Diabetes_Status, Sized by Age)") 
plt.tight_layout() 
plt.show()

Feature Correlation with Diabetes Outcome

corr_with_outcome = (df.corr(numeric_only=True)["Outcome"].drop("Outcome").sort_values(ascending=False))
corr_with_outcome.index.name = "" 

plt.figure(figsize=(8, 5))
sns.barplot(x=corr_with_outcome.values,y=corr_with_outcome.index,palette="magma")
plt.title("Feature Correlation with Diabetes Status", fontsize=14)
plt.xlabel("Correlation with Diabetes Status")
plt.tight_layout()
plt.show()

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import (confusion_matrix, classification_report,
                             accuracy_score, precision_score, recall_score,
                             f1_score, roc_auc_score, roc_curve)

from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier

Select Target Variable

df["Diabetes_Status"] = df["Diabetes_Status"].map({"No Diabetes": 0, "Diabetes": 1})
y = df["Diabetes_Status"]

Feature Selection

X = df.drop(columns=["Diabetes_Status", "Outcome", "Country"])

Encode Categorical Variables

from sklearn.preprocessing import OneHotEncoder

X = pd.get_dummies(X, drop_first=True)

Train-Test Split

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.2, random_state=42, stratify=y
)

Feature Scaling

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()

X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.transform(X_test)

from sklearn.impute import SimpleImputer

imputer = SimpleImputer(strategy="median")

X_train_imputed = imputer.fit_transform(X_train)
X_test_imputed = imputer.transform(X_test)

scaler = StandardScaler()

X_train_scaled = scaler.fit_transform(X_train_imputed)
X_test_scaled = scaler.transform(X_test_imputed)

log_model = LogisticRegression(max_iter=1000)
log_model.fit(X_train_scaled, y_train)

y_pred_log = log_model.predict(X_test_scaled)
y_prob_log = log_model.predict_proba(X_test_scaled)[:, 1]

cm_log = confusion_matrix(y_test, y_pred_log)

plt.figure(figsize=(5,4))
sns.heatmap(cm_log, annot=True, fmt="d", cmap="Blues")
plt.title("Confusion Matrix - Logistic Regression")
plt.xlabel("Predicted")
plt.ylabel("Actual")
plt.show()

print("Logistic Regression Classification Report:\n")
print(classification_report(y_test, y_pred_log))

print("Accuracy:", accuracy_score(y_test, y_pred_log))
print("Precision:", precision_score(y_test, y_pred_log))
print("Recall:", recall_score(y_test, y_pred_log))
print("F1 Score:", f1_score(y_test, y_pred_log))
print("ROC AUC:", roc_auc_score(y_test, y_prob_log))

Logistic Regression Classification Report:

              precision    recall  f1-score   support

           0       0.71      1.00      0.83      2569
           1       0.00      0.00      0.00      1031

    accuracy                           0.71      3600
   macro avg       0.36      0.50      0.42      3600
weighted avg       0.51      0.71      0.59      3600

Accuracy: 0.7136111111111111
Precision: 0.0
Recall: 0.0
F1 Score: 0.0
ROC AUC: 0.4927285296335212

DecisionTree Classifier

from sklearn.tree import DecisionTreeClassifier

dt = DecisionTreeClassifier(max_depth=6, random_state=42)
dt.fit(X_train, y_train)

y_pred_dt = dt.predict(X_test)
print(classification_report(y_test, y_pred_dt))

              precision    recall  f1-score   support

           0       0.71      0.99      0.83      2569
           1       0.24      0.01      0.02      1031

    accuracy                           0.71      3600
   macro avg       0.48      0.50      0.42      3600
weighted avg       0.58      0.71      0.60      3600

cm_dt = confusion_matrix(y_test, y_pred_dt)

plt.figure(figsize=(5,4))
sns.heatmap(cm_dt, annot=True, fmt="d", cmap="Greens")
plt.title("Confusion Matrix - Decision Tree")
plt.xlabel("Predicted")
plt.ylabel("Actual")
plt.show()

print("Decision Tree Classification Report:\n")
print(classification_report(y_test, y_pred_dt))

print("Accuracy:", accuracy_score(y_test, y_pred_dt))
print("Precision:", precision_score(y_test, y_pred_dt))
print("Recall:", recall_score(y_test, y_pred_dt))
print("F1 Score:", f1_score(y_test, y_pred_dt))
print("ROC AUC:", roc_auc_score(y_test, y_pred_dt))

Decision Tree Classification Report:

              precision    recall  f1-score   support

           0       0.71      0.99      0.83      2569
           1       0.24      0.01      0.02      1031

    accuracy                           0.71      3600
   macro avg       0.48      0.50      0.42      3600
weighted avg       0.58      0.71      0.60      3600

Accuracy: 0.7088888888888889
Precision: 0.24242424242424243
Recall: 0.007759456838021339
F1 Score: 0.015037593984962405
ROC AUC: 0.49901402191842675

RandomForestClassifier

from sklearn.ensemble import RandomForestClassifier

rf = RandomForestClassifier(
    n_estimators=200,
    max_depth=10,
    random_state=42
)

rf.fit(X_train, y_train)

y_pred_rf = rf.predict(X_test)
print(classification_report(y_test, y_pred_rf))

              precision    recall  f1-score   support

           0       0.71      1.00      0.83      2569
           1       0.00      0.00      0.00      1031

    accuracy                           0.71      3600
   macro avg       0.36      0.50      0.42      3600
weighted avg       0.51      0.71      0.59      3600

cm_rf = confusion_matrix(y_test, y_pred_rf)

plt.figure(figsize=(5,4))
sns.heatmap(cm_rf, annot=True, fmt="d", cmap="Oranges")
plt.title("Confusion Matrix - Random Forest")
plt.xlabel("Predicted")
plt.ylabel("Actual")
plt.show()

print("Random Forest Classification Report:\n")
print(classification_report(y_test, y_pred_rf))

print("Accuracy:", accuracy_score(y_test, y_pred_rf))
print("Precision:", precision_score(y_test, y_pred_rf))
print("Recall:", recall_score(y_test, y_pred_rf))
print("F1 Score:", f1_score(y_test, y_pred_rf))
print("ROC AUC:", roc_auc_score(y_test, y_pred_rf))

Random Forest Classification Report:

              precision    recall  f1-score   support

           0       0.71      1.00      0.83      2569
           1       0.00      0.00      0.00      1031

    accuracy                           0.71      3600
   macro avg       0.36      0.50      0.42      3600
weighted avg       0.51      0.71      0.59      3600

Accuracy: 0.7136111111111111
Precision: 0.0
Recall: 0.0
F1 Score: 0.0
ROC AUC: 0.5

Model Evaluation Metrics

from sklearn.metrics import roc_auc_score

print("ROC AUC:", roc_auc_score(y_test, rf.predict_proba(X_test)[:,1]))

ROC AUC: 0.4977764051650678

fpr_log, tpr_log, _ = roc_curve(y_test, y_pred_log)
fpr_dt, tpr_dt, _ = roc_curve(y_test, y_pred_dt)
fpr_rf, tpr_rf, _ = roc_curve(y_test, y_pred_rf)

plt.figure(figsize=(8,6))
plt.plot(fpr_log, tpr_log, label="Logistic Regression")
plt.plot(fpr_dt, tpr_dt, label="Decision Tree")
plt.plot(fpr_rf, tpr_rf, label="Random Forest")
plt.plot([0,1], [0,1], linestyle="--")

plt.title("ROC Curve Comparison")
plt.xlabel("False Positive Rate")
plt.ylabel("True Positive Rate")
plt.legend()
plt.show()

results = pd.DataFrame({
    "Model": ["Logistic Regression", "Decision Tree", "Random Forest"],
    "Accuracy": [
        accuracy_score(y_test, y_pred_log),
        accuracy_score(y_test, y_pred_dt),
        accuracy_score(y_test, y_pred_rf)
    ],
    "Precision": [
        precision_score(y_test, y_pred_log),
        precision_score(y_test, y_pred_dt),
        precision_score(y_test, y_pred_rf)
    ],
    "Recall": [
        recall_score(y_test, y_pred_log),
        recall_score(y_test, y_pred_dt),
        recall_score(y_test, y_pred_rf)
    ],
    "F1 Score": [
        f1_score(y_test, y_pred_log),
        f1_score(y_test, y_pred_dt),
        f1_score(y_test, y_pred_rf)
    ],
    "ROC AUC": [
        roc_auc_score(y_test, y_prob_log),
        roc_auc_score(y_test, y_pred_dt),
        roc_auc_score(y_test, y_pred_rf)
    ]
})

print(results.sort_values(by="ROC AUC", ascending=False))

                 Model  Accuracy  Precision    Recall  F1 Score   ROC AUC
2        Random Forest  0.713611   0.000000  0.000000  0.000000  0.500000
1        Decision Tree  0.708889   0.242424  0.007759  0.015038  0.499014
0  Logistic Regression  0.713611   0.000000  0.000000  0.000000  0.492729

from sklearn.ensemble import RandomForestClassifier

rf_model = RandomForestClassifier(random_state=42)
rf_model.fit(X_train, y_train)

RandomForestClassifier(random_state=42)

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

importances = rf_model.feature_importances_
feature_names = X.columns

feat_imp = pd.DataFrame({
    "Feature": feature_names,
    "Importance": importances
}).sort_values(by="Importance", ascending=False).head(15)

plt.figure(figsize=(8,6))
sns.barplot(data=feat_imp, x="Importance", y="Feature")
plt.title("Top 15 Feature Importances - Random Forest")
plt.show()